Chromosome segregation is an essential genetic function of eukaryotic cells. A critical component of the chromosome segregational machinery is the kinetochore, the centromere (CEN)- associated organelle at which spindle microtubules attach to the chromosome. A key event in kinetochore assembly is the packaging of CEN DNA into specialized nucleosomes containing a CEN-specific histone H3-like protein. Centric histone H3-like proteins are conserved from S. cerevisiae to humans, reflecting their fundamental importance in CEN function. The goal of this project is to elucidate the molecular mechanisms by which centromeric H3- like proteins are targeted to CEN DNA and assembled into CEN chromatin. The central hypothesis is that the specificity of centric H3-like proteins for centromeres is conferred by the CEN DNA itself, either (in cis) by its inherent structure (sequence, bends, flexibility), or (in trans) via nonhistone proteins specifically associated with it. I propose to test this hypothesis using the S. cerevisiae (yeast) model system. The yeast centromere H3-like molecule Cse4p and the other yeast core histones will be produced in E. coli, purified, reconstituted into histone octamers, and used to reconstitute nucleosomes on yeast CEN DNA in vitro. Binding preference for CEN DNA will be assessed using a competitive reconstitution assay in the presence and absence of purified CEN DNA binding proteins CP1, Mif2p, and Cbf3. If CEN specificity is observed, the critical cis-acting recognition sequences will be identified. Nucleosomes will also be reconstituted using Cse4p proteins lacking the essential N terminus to determine if the N terminus is required for CEN specificity or determining translational position. It is expected that the AT-rich CDEII element of the yeast centromere will be required for Cse4p nucleosome assembly, and genetic screens of random-sequence CDEII libraries will be carried out to determine the "sequence rules" for an optimal CDEII sequence. A Cse4p targeting assay will be used to determine which, if any, known yeast kinetochore proteins are required for targeting the centric histone to CEN DNA in vivo, and as a complementary approach, a genetic screen will be carried out to identify general Cse4p assembly factors. Given the high conservation of centric histone H3 variants throughout evolution, the results will illuminate the problem of centromere formation and function in higher eukaryotes including humans.